Statistics 550    Spring 2011    Exam 2     Take-home questions

1. Exercise 12.4 in the text.

2. Exercise 12.10 in the text.

3. For the Spokane houses data set, develop a regression model to predict house price (recorded in thousands of dollars) from the other variables that were collected.  The other variables were square footage of the house, number of bedrooms, number of bathrooms, and whether the house was a single family residence or had land included with the house.  Use an iterative approach where diagnostic results from your first models lead to a final model, and show plots or report results that lead to model changes.  For your final model assess the assumptions of normality, constant variance, and linearity.  For your final model use a discrete-data-based test for nonlinearity, and report on whether any of the observations are unusual or influential.  Include a summary paragraph that describes your model-fitting process and reports on the results for your final model.